EnergyPlus Simulation Run Time Diagnostics

asked 2019-03-07 10:13:23 -0600

jt1234
331 ●1 ●3

I realize there are lots of posts and documents related to EPlus simulation time and solutions for mitigating long run time. My question takes these posts/solutions a step further and looks at what exactly is causing the slow down. I've had scenarios where I inadvertently figure out a particular zone/object that was causing a slow down - but entirely by dumb luck. Is there any way to dissect or expand the simulation run log in order to figure out if-or-which zone/object is causing the slow-up or getting stuck in a calculation loop?

This may not be possible, but worth throwing out to the group.

answered 2019-03-07 11:28:14 -0600

mdahlhausen

9499 ●19 ●11 http://openstudio.net/

Your intuition to check the run log and .err file is correct. Some routines may take a while to converge because of bad inputs, and the run log and .err file will usually show which zone and components are responsible. Unfortunately there isn't a way to directly determine which routines are causing a slow down. That is part of a current DOE BTO funded project EnergyPlus 10X to speed-up simulation run time by a factor of 10. Likely through refactoring the surface/zone heat balance routines, HVAC loop simulation, and parallelization.

Until that work is complete, here are some things known to cause high run times (an incomplete list):

Lots of zones
Lots of surfaces
Plenums that connect to lots of surfaces / zones
Stratified water heaters
Individual air loop / plant loops / zone equipment for each zone in a multi-zone model
Lots of shading and high resolution shading/daylighting calculations
20+ design days